Multiple Language Database Design - Additional Columns vs. Additional Tables...

  • Gift Peddie

    SSC Guru

    Points: 73570

    This thread is actually about doing that only you need three columns because Arabic is right to left language, Korean does not use the Latin alphabet it comes with the more than 2000 character east Asian languages alphabet and English uses the Latin alphabets. The reason we are counting bytes and all three falls under three separate bytes ranges.

    Kind regards,
    Gift Peddie

  • tymberwyld

    SSCertifiable

    Points: 7810

    Most applications attempt to do this by having a single "NVARCHAR" column for FirstName, LastName, etc. This post is just about storing that information in a more efficient manner. You could probably attempt the single NVARHCAR column for you database at first. I think this post is starting to blend into content management for multi-lingural sites as well ;-).

  • Adrian Robertson

    Mr or Mrs. 500

    Points: 531

    Thanks

    I am able to store both Korean text and English in the same field (either for different records or for the same record) but the way I achieved this was by using the Korean_Wansung collation on the DB itself, rather than Latin_General

    Just not sure whether this would then allow me to store other languages at the same time (such as Japanese or Thai), though I'm guessing not, especially when it comes to Thai as it's a more complex character set

    So would it be fair to say that the collation can be set at a field level, rather than at a DB level?

  • tymberwyld

    SSCertifiable

    Points: 7810

    Yes. I'm not sure what the best approach is (hence why we have this thread going). You may be able to put your database back to "latin" and then add seperate columns for each language outside the "latin" scope. For example, you should be able to store English, Portuguese, Spanish (all latin languages) in one NVARCHAR column, then add others for Chinese, Korean, Japanese, etc. This thread was an attempt to make that all kind of transparent without having to add more columns and more complication over time as you want to support new languages.

  • Gift Peddie

    SSC Guru

    Points: 73570

    This thread was an attempt to make that all kind of transparent without having to add more columns and more complication over time as you want to support new languages.

    I am not sure I agree on the above point because yes Korean falls under the same alphabet scheme as Japanese but one version of Japanese is very hard to store so it depends and Thai is actually a right to left language like Arabic so it is not related to Korean.

    Kind regards,
    Gift Peddie

  • Gift Peddie

    SSC Guru

    Points: 73570

    would not be so sure about that.

    They issued a patent to Microsoft for the Cascading Style Sheets.

    Microsoft is the victim of a frivolous data patent so I am hoping Redmond will help persuade the US Government from issuing patents that did not provide value but impede innovation.

    http://news.yahoo.com/s/nf/20090819/bs_nf/68501

    😉

    :hehe:

    Kind regards,
    Gift Peddie

  • JacekO

    SSCertifiable

    Points: 6347

    Gift,

    You just blew my mind off...

    AMAZING.

    It is not about the article (which is quite interesting in itself) but the fact you remembered about an obscure message posted several months ago related to this article...

    I have hard time remembering what I had for breakfast yesterday...

    ---------------------------------------------
    [font="Verdana"]Nothing is impossible.
    It is just a matter of time and money.[/font]

  • Gift Peddie

    SSC Guru

    Points: 73570

    JacekO (8/20/2009)


    Gift,

    You just blew my mind off...

    AMAZING.

    It is not about the article (which is quite interesting in itself) but the fact you remembered about an obscure message posted several months ago related to this article...

    I have hard time remembering what I had for breakfast yesterday...

    Hi Jaceko,

    Globalization posts are not obscure to me in most cases because I think you can develop in English add translated static content and use the same application in many languages as needed. If this thread is about Backup or other areas of SQL Server then it will be obscure to me.

    Kind regards,
    Gift Peddie

  • nasimhmd7

    Valued Member

    Points: 71

    I think I was having problems with compatibility, so decided to unistall-reinstall, now, I can't get VS2008 to load an sdf file.

    Natural Colon Cleanse [/url]

  • sam.walker

    Mr or Mrs. 500

    Points: 597

    I might be shunned from the relational community, for making the mere point.

    But if there is technology in place, which permits the storage/extraction of different language versions for a bit of text within a single field, and where we won't have to create these extra tables.

    Is there any harm in encoding the data in to a single field, until such technology becomes available?

    And could one way of doing this be with a tightly constrained XML field?

    Sam

    <Text>

    <Translation locale="en-US" isDefault="True">G'day</Translation>

    <Translation locale="en-GB">Good day to you</Translation>

    <Translation locale="fr-FR">Bonjour</Translation>

    </Text>

  • aspcreator

    SSC Rookie

    Points: 44

    We have our own solving - special framework: Digital Data Cube (www.digitaldata3d.eu). This framework is focused on development of large database applications and implements many features. For example:

    - simultaneously multi languages of data - with one database may work a lot of different users around of the world with different languages. Data are retrieved for required language of each user (must be translated). Count of used languages is not limited.

    - Full history of change data - each change may be possible approved or rejected up to 49 levels. 50 level - change is approved.

    - Implementation of the maker checker - maker is user who create record, checker is user who changed or approved the change of the record. Each of them has access rights and may be defined that checker can not approve the changes himself.

    - Notes - to each record may be attached text notes (0 or more)

    - Attachments - to each record may be attached binary attachments (0 or more) like is pdf, excel, images and so on.

    - framework is prepared for multi-user environment - changes are protected before simultaneously editation of another user

    and more and more other specific behaviours.

    After training which is required is using of Digital Data Cube very easy with maximum elimination of errors in the source code.

    Now we are redesigning our web sites for other language (En, Ge) for easily accessibility of other candidates. For more detail information, please, contact us on info@digitaldata3d.eu

Viewing 11 posts - 31 through 41 (of 41 total)

You must be logged in to reply to this topic. Login to reply